Procedure - Participants were shown 15lists each containing 15 words. They were to recall them immediately after being shown the list, recall after a 10 second interference task (which involved countingbackwards from a number in threes), and recall after a 30 second interference task.
Glanzer & Gunitz 1966
Results - Participants in the immediaterecall condition showed the serialpositioneffect (primary is more items recalled from beginning and recency is end of list). 10 second delay showed a similar primary effect but less recency. 30 second delay had highest primary effect but no recency effect
Glanzer & Gunitz 1966
Conclusion - The 30 second interference task prevented rehearsal, so items at the end of the list couldn't be transferred to LTM. So, there appears to be twoseparate storage facilities.